Abstract Background: Whether feature or spatial-temporal information is critical for maintenance of persisting object representation is still debated. The similarity theory emphasizes the role of feature information. However, object file theory proposes the spatial-temporal continuity of object position help maintain the persisting object representation. Previous studies based on the object file framework revealed that feature information alone could not support persisting object representation. However, those studies neglected the difference between the static feature and the feature change. The present study tested the hypothesis that the continuity of feature change may be a necessary condition for maintenance of persisting object representation. Method: The object reviewing paradigm was adopted. In each trial, two characters flashed briefly in two circles. Then the circles moved smoothly to vertical positions. One probe character appeared in one of circles at the end of moving. Participants made a rapid judgment on whether the character was one of the two characters appeared at the beginning of moving. If the probe was one of two characters presented before, it might appear in the congruent circle (congruent condition) or in the incongruent circle (incongruent condition). Object Specific Preview Benefits (OSPB), the difference between reaction times of the incongruent condition and the congruent condition, was taken as an index of how well the persisting object representation was preserved. The continuity of feature change was manipulated as a between subject factor within each experiment. There were 3 levels of continuity: No-change, Gradual-change and Abrupt-change (or Random-change). Experiment 1 tested whether shape change continuity affected object persistence. In the No-Change condition (n = 10), the object shape did not change during the moving. In the Gradual-Change condition (n = 17), the object shape changed gradually from a circle to an oval shape during moving. In the Abrupt-Change condition (n = 11), the object shape transformed between a circle and an oval shape during moving. Experiment 2 tested whether brightness change continuity affected object persistence. In the No-Change condition (n = 9), the object brightness was constant during moving. In the Gradual-Change condition (n = 12), the object brightness changed gradually during moving. In the Random-Change condition (n = 6), the object brightness changed randomly during moving. Results: In Experiment 1, significant OSPB was found in the no-change condition (OSPB = 28.32ms, p = .006) and the gradual-change condition (OSPB = 29.31ms, p = .004), but not in the abrupt-change condition (OSPB = -2.57ms, p = .82). Similar results pattern was also found in Experiment 2. Significant OSPB was found when brightness did not change (OSPB = 78.26ms, p = .001) and when brightness changed gradually (OSPB = 26.74ms, p = .015), but was not found when brightness change randomly (OSPB = 1.88ms, p = .934). Conclusion: These results indicate that the continuity of feature change is also necessary for the maintenance of persisting object representations. The present study implies that object file theory may need to include both the continuity of feature change and the continuity of position change to define the spatial temporal continuity.